PROVINCIAL SUPERVISION EXPLORATION OF HISTORIC CITIES, TOWNS AND VILLAGES IN CHINA BASED ON DEEP LEARNING AND GIS – TAKE ZHEJIANG PROVINCE AS AN EXAMPLE

Abstract. At present, China is building a historic heritage conservation and management system from upper to lower level. Therefore, as the middle level, provincial level is becoming increasingly important in the management of famous towns and villages. However, due to the excessive quantities of resources, the lagging management skills and other influential factors, the provincial competent authorities still face a series of difficulties in implementing the conservation and management of historic cities, towns and villages. Focusing on the regulatory issues faced by provincial competent authorities, this study combined geospatial information technology with deep learning technology to explore the intelligent auxiliary identification of traditional texture for historic settlements and the provincial planning guidance method based on GIS evaluation. This paper put forward technical optimization suggestions for the stages of declaration and identification, conservation planning, supervision and inspection in provincial conservation and management. The intervention of new technologies provides a reference for provincial competent authorities in enhance conservation and regulatory capacity, and dealing with the situation of "having laws but lacking technology, and difficult to implement".